With respect to optical sensors of the related art, a device which is prepared by forming a photodiode (PD) in a semiconductor substrate such as silicon (Si) is general. As a solid-state imaging device, there is widely employed a planar solid-state imaging device in which PDs are two-dimensionally arranged in a semiconductor substrate and a signal corresponding to a signal charge as generated from each PD by photoelectric conversion is read out by a CCD or CMOS circuit. As a method of realizing a color solid-state imaging device, a structure in which color filters each of which is able to transmit only light having a specific wavelength therethrough are arranged for the color separation in a side of the light incident face of a planar solid-state imaging device is general. In particular, as a system which is widely employed at present for digital cameras and so on, there is well known a single solid-state imaging device in which color filters which are able to transmit blue (B) light, green (G) light and red (R) light, respectively therethrough are regularly arranged on respective two-dimensionally arranged PDs.
However, in the single solid-state imaging device, since the color filter transmits only light of a limited wavelength therethrough, light which has not transmitted through the color filter is not utilized, and the use efficiency of light is poor. Also, with progress of high integration of pixel, the size of PD becomes the same in size as the wavelength of light, whereby the light is hardly waveguided into PD. Also, since blue light, green light and red light are detected by separate PDs adjacent to each other and then subjected to arithmetic processing, thereby achieving color reproduction, a false color may possibly be generated. In order to avoid this false color, an optical low-pass filter is necessary, resulting in the generation of an optical loss by this filter.
There have hitherto been reported color sensors in which three PDs are stacked within a silicon substrate by utilizing the wavelength dependency of an absorption coefficient of silicon and color separation is carried out due to a difference in depth on the pn junction face of each PD (U.S. Pat. Nos. 5,965,875, 6,632,701 and Jp-A-7-38136). However, such a system involves a problem that the wavelength dependency of spectral sensitivity in the stacked PDs is broad so that the color separation is insufficient. In particular, the color separation between blue and green colors is insufficient.
In order to solve this problem, there is proposed a sensor in which a photoelectric conversion part which detects green light and generates a signal charge corresponding thereto is provided in an upper part of a silicon substrate and blue light and red light are detected by two PDs as stacked within the silicon substrate (see Jp-A-2003-332551). The photoelectric conversion part which is provided in the upper part of the silicon substrate is configured to include a first electrode layer as stacked on the silicon substrate, a photoelectric conversion layer which is made of an organic material as stacked on the first electrode layer and a second electrode layer as stacked on the photoelectric conversion layer. This photoelectric conversion part is configured such that when a voltage is applied to the first electrode layer and the second electrode layer, a signal charge as generated within the photoelectric conversion layer transfers into the first electrode layer and the second electrode layer and a signal corresponding to the signal charge as transferred into either one of the electrode layers is read out by a CCD or CMOS circuit or the like as provided within the silicon substrate. In this specification, the “photoelectric conversion layer” as referred to herein means a layer capable of absorbing incident light having a specific wavelength which has been made incident thereinto and generating an electron and a hole corresponding to the quantity of absorbed light.